Whole genome bisulfite sequencing (WGBS) and microarray studies have linked aberrant methylation patterns to a variety of cancers. However, WGBS is costly and includes coverage of uninformative regions, and microarrays only detect a small fraction of all CpG sites. To minimize cost and enable thorough methylome-level measurements at scale, Twist Bioscience has designed a solution to enrich the most current, annotated, and relevant CpG sites within the human genome.

The Twist Human Methylome Panel, in combination with the Twist Methylation Detection System, is an optimized end-to-end workflow designed to target more than 3.98 million biologically relevant CpG sites, with an additional 2.61 million observed when shadow coverage is included. Here, we present these products as an end-to-end solution. The Twist Human Methylome Panel spans 123Mb of total probe space of human enhancers, promoters, transcription factor binding sites, and CpG islands, curated from the most recent database releases. Twist’s panel design approach uses predictive modeling to reduce issues that arise due to a lack of sequence complexity related to the conversion process in methyl-seq applications. Using pre-capture enzymatic conversion technology demonstrated by Twist’s system, DNA damage is minimized and library mass is increased. Further, the Twist Methylation Detection System is designed to enable tuning of protocol conditions to ensure flexibility in different contexts.

At 150x raw coverage, the Twist end-to-end solution achieved a Fold-80 Base Penalty of 1.54; Percent Target Bases at 30X of 90%; and Percent Off-Bait of 5%. 6.59 million CpG sites were detected using a minimum depth of 10X, representing 23% of the available CpG sites found in the human genome. This compares favorably to leading methylation microarray-based methods, which only measures ~850k CpG sites. Finally, the BWA-meth/MethylDackel workflow was found to be the most robust method to align reads, handle the presence of SNPs, and produce high quality CpG calls.

Our study demonstrates an efficient end-to-end method to interrogate genome-wide methylation patterns for various applications. Not only does using this target enrichment system decrease the cost of methylation detection compared to WGBS, but it also provides higher sensitivity for differential methylation detection compared to microarray technology.